Nicotinamidephosphoribosyltransferase and cerebral ischemia

1. Introduction

Stroke is the second most common cause of death worldwide and the leading cause of long-term adult disability in the United States. Thus far, the only drug that has been used successfully to treat acute stroke in the clinic is the clot-dissolving drug tissue plasminogen activator (tPA). However, this medication is restricted in use because of the limited temporal window (within three hours of stroke symptom onset) during which it can be safely applied [1]. To date, numerous cytoprotective drugs have reached phase 3 efficacy trials in acute stroke patients. Unfortunately, throughout the clinical literature, the phrase “failure to demonstrate efficacy” prevails as a common thread, despite mostly encouraging results in preclinical studies (Labiche LA 2004 46). It is therefore imperative to develop new neuroprotective and neurorestorative strategies for ischemic stroke. Endogenous defense mechanisms against stroke injury may hold the key to the development of novel therapies for stroke. A large number of studies have suggested that nicotinamidephosphoribosyl-transferase (NAMPT) is an attractive candidate to improve post-stroke recovery. NAMPT is the rate-limiting enzyme in the salvage pathway of nicotinamide adenine dinucleotide (NAD) biosynthesis. Depletion of NAD due to energy depletion and PARP-1 consumption is a critical step in the induction of cell death after ischemia [2–4]. Thus, as the key regulator of NAD biosynthesis, NAMPT has been postulated to exert a significant impact on cellular function and survival under ischemic conditions. Besides regulating energy metabolism, recent studies have confirmed that NAMPT has multiple functions: enhancing the proliferation of endothelial cells, inhibiting apoptosis, regulating vascular tone, and stimulating autophagy. Lim et al. have shown that the administration of NAMPT recombinant protein at the time of reperfusion reduces infarct size by approximately 20% in a mouse model of cardiac ischemia [5]. Genetic deletion or pharmacological inhibition of NAMPT worsens ischemic brain injury [6] and lentivirus-mediated NAMPT overexpression protects against ischemic brain injury [7]. The precise neuroprotective mechanisms underlying the efficacy of NAMPT in cerebral ischemia remain unknown, although previous studies have suggested that NAMPT may mediate neuroprotection by providing extra energy and activating NAD-dependent survival signaling pathways. These studies provide rationale for investigating NAMPT in ischemic stroke. However, there is little agreement on the effects of NAMPT on ischemic injury. Some reports have argued that NAMPT exacerbates ischemic neuronal injury non-enzymatically by triggering the release of tumor necrosis factor alpha (TNF-α) from glial cells [8]. In addition, NAMPT also participates in several pathophysiological processes such as hypertension, atherosclerosis, and ischemic heart disease. Thus, the conditions under which NAMPT is beneficial remain unclear. Recent studies using in vitro and in vivo genetic/pharmacologic manipulations, including our own studies, have shed some light on the answer to this question. This review focuses on 1) basic characteristics of NAMPT that have the potential to be modified in pathological conditions, 2) various complex functions of NAMPT that tie in to ischemic conditions, and 3) the underlying mechanisms.

2. Basic characteristics of NAMPT

The human NAMPT (hNAMPT) gene (Gene ID: 10135) has 37,643 bp and is located on chromosome 7q22.3 according to Pubmed (2014). The structural NAMPT gene is composed of 11 exons and 10 introns. The regulatory gene at the 5′-flanking region has a 1.4 kb region that is GC rich (60% GC) following a 1.6 kb region with more AT bases (60% AT). The first 500 bp upstream of the major transcription initiation site (TIS) lacks canonical TATA and CAAT boxes but has higher G+C content (72% GC). Several CAAT boxes and TATA-like and initiator sequences are found about 2 kb upstream from the TIS, which may act as the promoter region. The binding sites for transcription factors such as nuclear factor 1 (NF1) and selective promoter factor 1 (Sp1) activator protein (AP-1/AP-2) and hepatocyte nuclear factors (HNFs) have been identified. The binding sites for the glucocorticoid receptor, corticotropin releasing factor, cAMP response element binding protein, and the nuclear factors (such as NF-IL6, NF-κB) are found in the promoter region. These multiple regulatory elements hint at diverse means of transcriptional modulation [9].

By December 24, 2014, 1480 SNPs in the hNAMPT gene have been reported. The functional consequences of most of these SNPs are currently unknown. In general, gene function may be affected when SNPs occur within a regulatory region near a gene [10]. Ye et al. discovered a 7.7-fold higher risk of acute respiratory distress syndrome (ARDS) with two SNPs (T-1001G and C-1543T) in the hNAMPT gene promoter, and TT represents a protective haplotype against ARDS [10]. Bajwa et al. found that the NAMPT T-1001G variant allele and related haplotype are associated with increased odds of developing ARDS and increased intensive care unit mortality among at-risk patients, whereas the C-1543T variant allele and related haplotype are associated with decreased odds of ARDS among patients with septic shock and better outcomes among patients with ARDS [11]. The G-948T gene polymorphism was associated with increased high-density lipoprotein (HDL) cholesterol in obese subjects [12]. The −1535T/T genotype was associated with lower serum triglyceride levels and higher HDL-cholesterol levels in non-diabetic Japanese subjects [13]. rs9770242 and rs1319501 were in perfect linkage disequilibrium, whereas fasting insulin and glucose levels and rs7789066 were significantly associated with the apolipoprotein B component of very low density lipoprotein (VLDL) from the Quebec Family Study [14]. rs9770242, −948G-T, and rs4730153 were associated with the ratio of visceral/subcutaneous visfatin mRNA expression [15]. In addition, −948G-T was associated with 2-h plasma glucose and fasting insulin concentrations in nondiabetic Caucasian subjects [15]. These studies reveal that the NAMPT gene may play an important role in metabolic function.

For the hNAMPT gene, transcription produces 19 different mRNAs, 14 alternatively spliced variants and 5 unspliced forms. There are 5 probable alternative promoters, 6 non-overlapping alternative last exons and 13 alternative polyadenylation sites. The mRNAs appear to differ by truncation of the 5′ end, truncation of the 3′ end, presence or absence of 2 cassette exons, overlapping exons with different boundaries, alternative splicing or retention of 4 introns. The physiological relevance of these NAMPT variants are currently unknown, and assumed to be even more complicated under pathological conditions.

NAMPT mRNA is ubiquitously distributed in all tissues including heart, brain, placenta, lungs, liver, skeletal muscle, kidney blood vessels, and pancreases. Liver and muscle tissue show the highest mRNA levels [9,16–18].

NAMPT protein is highly conserved from invertebrates to mammals, including humans. The canine NAMPT protein sequence is 96% identical to the human sequence and 94% identical to murine and rat counterparts [19]. The mammalian gene is functionally active in prokaryotic hosts [20]. The hNAMPT protein sequence (NP_005737.1) consists of 491 amino acids with a predict molecular weight of 52 kDa. It contains 63 negatively charged residues (Asp + Glu) and 61 positively charged residues (Arg + Lys). hNAMPT is considered to be a stable protein, since the estimated half-life is 30 hours based on the methionine at the N-terminal of the sequence.

Several groups have resolved the three dimensional structure of mammalian NAMPT protein [21]. The structure of the NAMPT monomer has three domains containing total 22 β-strands and 15 α-helices. Homodimerized NAMPT catalyzes the conversion of nicotinamide (NAM) and phosphoribosyl-pyrophosphates to form nicotinoamide mononucleotide, a key step in the mammalian NAD synthetic salvage pathway. The 3D structure analysis of human and murine NAMPT by Khan et al. [22] showed that NAMPT is functionally similar to quinolinic acid and nicotinic acid phosphoribosyltransferases, enzymes in two other NAD biosynthesis pathways, despite having la arger size and highly divergent amino acid sequences.

NAMPT catalyzes ATP hydrolysis and become autophosphorylated at a histidine residue (His247) that lies at the center of a conserved cluster of active site residues. Autophosphorylation can increase its enzymatic activity and lower the Km for the substrates [23]. Wang et al reported that NAMPT is in the dimeric class of type II phosphoribosyltransferases with a unique hydrogen bond between Asp219 and the amide of nicotinamide. Mutations of the residue H247 significantly affect NAMPT enzymatic activity [24]. The structural and mutagenesis studies by Khan et al. [16] demonstrate that Asp219 is important in defining the substrate specificity of NAMPT. The structural differences between these enzymes indicate variations in substrate specificity and inhibitor sensitivity. Furthermore, although the 5′-flanking region of the gene of NAMPT lacks the classical secretion signal sequences [9], it can be secreted via an endoplasmic reticulum–Golgi or microvesicles independent pathway [25,26].

3. Regulation of NAMPT

The basic characteristics and regulation of NAMPT have been revealed by a plethora of studies using small molecule inhibitors, siRNA, miRNA, antisense oligonucleotide, gene overexpression, gene knock out, and antibodies. The current chemical inhibitors of NAMPT are analogues of FK866 (APO866, (E)-N-[4-(1-benzoylpiperidin-4-yl) butyl] acrylamide-3-(pyridin-3-yl) and CHS 828 (N-(6-chlorophenoxyhexyl)-Nt-cyano-Nu-4-pyridylguanidine).

The use of FK866 as a NAMPT inhibitor was first reported by Hasmann and Schemainda [27]. FK866 induced apoptosis in HepG2 human liver carcinoma cells by a highly specific and potent inhibition of NAMPT with an IC₅₀ of approximately 1 nM [27]. FK866 did not directly inhibit mitochondrial respiratory activity, but caused gradual NAD⁺ depletion by inhibiting NAMPT, which eventually triggers apoptosis. Nicotinic acid and NAM were both found mitigate the cellular effects of FK866. The mechanism underlying FK866 inhibition of NAMPT has been well characterized. The crystal structure of NAMPT, free and in complex with FK866, has been identified by Wang et al [24] and Kim et al [28]. FK866 binds at the nicotinamide-binding site of NAMPT (a tunnel at the interface of the NAMPT dimer), thus functioning as a competitive inhibitor of NAMPT enzymatic activity [22,24,28]. Mutations in this binding site can abolish the inhibition by FK866. FK866 binds to the active site of NAMPT with higher affinity than either the substrate or the product. The benzoylpiperidin group of FK866 plays a key role in binding to NAMPT in hydrophobic interactions [28]. As NAMPT acts as a key enzyme of NAD synthesis, FK866 displays strong anticancer activity in hematologic malignancies both in vitro and in vivo [29,30].

CHS 828, a cyanoguanidine compound, is a structurally unique NAMPT inhibitor currently in phase II clinical trials for cancer patients [31–34]. CHS 828 inhibits NAD synthesis in a similar way as FK866. Increasing cellular levels of NAD can completely block the cellular effects of CHS 828, suggesting a competitive mode of action. Furthermore, the crystal structure and in vitro biochemistry shows that FK866 and CHS 828 share a common binding site at the active site of NAMPT [28,35,36].

As mentioned in Section 2 (Basic characteristics of NAMPT), NAMPT can be upregulated in several different ways. The activation of NAMPT by H247-phosphorylation causes stabilization of the enzyme-phosphoribosylpyrophosphate complex, permitting efficient capture of nicotinamide. Small changes in distances in the reaction coordinates also have large effects on catalysis, and may account for the activation of NAMPT when phosphorylated at H247. The activation of NAMPT by ATP hydrolysis also suggests that physiological enzymatic activation is a consequence of the formation of phospho-His-247 (Burgos ES 2008 11086) [24]. Recent studies have identified NAMPT as the intracellular target of P7C3, a proneurogenic compound containing an aminopropyl carbazole. P7C3 enhance the enzymatic activity of purified NAMPT by increasing NAD levels through NAMPT-mediated salvage pathways. These studies leave unanswered how exactly P7C3 works, as the native NAMPT enzyme is regulated by posttranslational modifications, intracellular localization, and a wide spectrum of other intracellular variables (Wang g 2014 1324).

4. Physiological functions of NAMPT

Accumulating evidence indicates that NAMPT has at least three major functions–as a nicotinamidephosphoribosyltransferase, a growth factor, and a cytokine.

5. Intracellular and extracellular forms of NAMPT

NAMPT exists in either the intracellular or extracellular space. iNAMPT was found in nuclei, mitochondria, and the cytosol [17]. iNAMPT participates as a rate-limiting enzyme in the salvage pathway of NAD synthesis and plays multiple roles in energy metabolism, enhancing the proliferation of endothelial cells, inhibiting apoptosis, regulating vascular tone, and stimulating autophagy in disease conditions such as stroke [51]. eNAMPT has been found both locally in non-CNS systems and circulating systemically in the blood. eNAMPT can be distinguished from iNAMPT by its larger molecular weight arising from posttranslational modifications [38]. eNAMPT, mostly in the form of serum NAMPT, likely functions as a cytokine within the peripheral circulation. eNAMPT can be released from adipocytes and participate in energy homeostasis [58]. A large number of studies have reported effects of eNAMPT on various biological functions. It is widely believed that the diverse biological functions of eNAMPT can be attributed to its NAMPT enzymatic activity. However, due to the scarcity of ATP in the extracellular space, the enzymatic activity of eNAMPT may be weak under normal circumstances [75].

6. Regional and cell-specific distribution of NAMPT in the brain

NAMPT exists ubiquitously in various tissues in human and murine brains [16,17,52]. McGlothlin et al described moderate expression of NAMPT in the brain [19]. In the study by Zhang and colleagues, the expression of NAMPT was shown to be almost exclusive to neurons in the mouse brain, absent in glial cells, and present in some endothelial cells [6]. Wang et al have shown NAMPT is expressed in both neurons and astrocytes in vitro. However, the expression of NAMPT in neurons was higher than in astrocytes, suggesting perhaps a more important role in the former cell type [7]. Liu et al showed that NAMPT levels increased in serum with aging but decreased in the aging brain in the cortex and hippocampus with no parallel change in striatum and cerebellum. Furthermore, NAMPT increased with aging in microglia but likely decreased in neurons. In young mice, immunoreactivity for NAMPT was primarily localized in NeuN-positive neurons in the cortex and hippocampal CA3 region but rarely in GFAP-positive astrocytes and Iba1-positive microglia cells. NAMPT was also found to be expressed in the Purkinje cells, granule cells, and cells in the molecular layer of cerebellum. In aged mice, NAMPT was also found to be expressed in neurons. In addition, NAMPT was highly expressed in microglial cells, but not in astrocytes [76].

When wild type hNAMPT or H247A mutant NAMPT with very weak enzymatic activity were injected intravenously into mice once every 3 d for 32 d, body weight, blood pressure, heart rate, serum glucose, serum total cholesterol and triglyceride and the morphology of neurons, astrocytes and microglia in the hippocampus were not significant affected [77]. This suggests that elevation of plasma NAMPT may not induce metabolic and neuronal dysfunction in normal individuals. In a focal cerebral infarction model, NAMPT expression was markedly increased in both the infarct core and peri-infarct penumbra, compared with the corresponding unaffected contralateral region. Additionally, NAMPT was upregulated in an in vitro model of ischemic neuronal injury from oxygen–glucose deprivation (OGD) [7].

7. Roles of NAMPT in brain ischemia

Compared with brain ischemia, the roles of NAMPT in the ischemic heart have been more widely studied. Recombinant NAMPT administration reduces myocardial injury both in vivo and in vitro, providing direct evidence for the cardioprotective effects of NAMPT. These effects were blocked by a phosphoinositide 3-kinase (PI3K) inhibitor and a MEK1/2 inhibitor [5]. Expression of NAMPT in the heart is decreased by ischemia/reperfusion and pressure overload. Cardiac-specific overexpression of NAMPT in mice increases NAD⁺ content in the heart, prevents downregulation of NAMPT, and reduces myocardial infarction size and apoptosis in response to prolonged ischemia and ischemia/reperfusion. These results suggest that NAMPT is an essential gatekeeper of energy status and survival in cardiac myocytes [78]. As downregulation of NAMPT impairs autophagic flux, Hsu et al suggested that NAMPT may protect against myocardial infarction by positively regulating autophagy in cardiomyocytes [78,79]. These experiments suggest that NAMPT may be an important pro-survival molecule in cardiomyocytes and that the increase of NAMPT during heart ischemia is a compensatory defense mechanism. There are many similarities between the pathophysiological mechanisms underlying ischemic heart disease and ischemic stroke. NAMPT may mediate neuroprotection by providing extra energy and activating NAD-dependent survival signaling pathways. These studies warrant further investigations of NAMPT in ischemic stroke.

As mentioned in the Introduction, stroke is a leading cause of morbidity and mortality worldwide. Ischemia leads to energy depletion and eventually neuronal death and brain damage [80,81]. NAMPT has been postulated to exert a significant impact on cellular function and survival under ischemic conditions by its multifaceted functions, including but not limited to energy metabolism. Below we review the recent literature on the complex effects of NAMPT on ischemic injury.

Lu et al reported that plasma NAMPT concentrations increased in ischemic stroke patients as an independent factor associated with ischemic stroke [82]. Similar observations were made in a mouse model [18,83]. Both eNAMPT and iNAMPT protein expression was detectable with two different NAMPT antibodies in WT C57BL/6 mice. Although iNAMPT remained unchanged, eNAMPT was dramatically induced at 24 hours after ischemia, peaking at 3 days and remaining at high levels even after 7 days. NAMPT expression was induced in both neuronal cell bodies and white matter (WM) regions, including striatal fiber bundles and the corpus callosum in WT mice after ischemia. These findings suggest that eNAMPT is selectively induced after ischemia, perhaps as part of a long-term stress response pathway.

Wang et al used lentivirus-mediated NAMPT overexpression and knockdown to manipulate NAMPT expression and explore the effects of NAMPT on neuronal survival during ischemic stress using in vivo middle cerebral artery occlusion and in vitro OGD models. Both local lentivirus-mediated NAMPT overexpression and administration of the NAMPT enzymatic product NMN reduced ischemia-induced cerebral injuries [7]. NAMPT inhibition by FK866 aggravated brain infarction after cerebral ischemia in rats. The neuroprotection exerted by NAMPT was abolished in adenosine monophosphate-activated kinase-[alpha]2 knockout (AMPK α2^−/−) neurons, which supports the notion that NAMPT overexpression and knockdown regulates neuronal survival by the AMPK pathway. In neurons, NAMPT positively modulates NAD⁺ levels and thereby controls SIRT1 activity. SIRT1 can be co-precipitated with serine/threonine kinase 11 (LKB1), an upstream kinase of AMPK, and has been shown to promote LKB1 deacetylation in neurons, thereby controlling activation of the AMPK signaling pathway. NAMPT overexpression-induced neuroprotection was abolished in SIRT1^+/− and AMPKα2^−/− mice. Therefore, SIRT1 appears to be essential for NAMPT-induced AMPK activation and neuroprotection. Thus, NAMPT protects against ischemic stroke by rescuing neurons from death via the SIRT1-dependent AMPK pathway [7]. Subsequent work demonstrated that the overexpression of NAMPT increased autophagy, as indicated by LC3 expression in punctate structures, LC3-II/beclin-1 expression and increased autophagosome number both in vivo and in vitro at 2 hours after ischemic injury. At the early stages of OGD, the autophagy inducer rapamycin protected against neuronal injury induced by NAMPT knockdown, whereas the autophagy inhibitor 3-methyladenine partially abolished the neuroprotective effects of NAMPT. Overexpression or knockdown of NAMPT regulated the phosphorylation of mTOR and S6K1 signaling pathways after OGD stress by enhancing phosphorylation of TSC2 at Ser1387 but not Thr1462. Furthermore, in cultured SIRT1-knockout neurons, the regulation of autophagic proteins LC3-II and beclin-1 by NAMPT was abolished. These results reveal that NAMPT promotes neuronal by promoting autophagy through the TSC2-mTOR-S6K1 signaling pathway in a SIRT1-dependent manner during cerebral ischemia.

Heterozygote NAMPT-knockout (NAMPT^+/−) mice with reduced expression of NAMPT have been used to determine the role of NAMPT in cerebral ischemia induced by photothrombosis. NAMPT^+/− mice exhibit a higher density of degenerating, Fluoro-Jade B+ neurons in the penumbra than WT mice. These results provide evidence that neuronal NAMPT is protective against cerebral ischemia [6]. In vitro OGD and glutamate excitotoxicity models have been used to study the mechanism underlying the protective effects of NAMPT on primary neurons. Treatment of primary neurons with NAM and NAD⁺ significantly reduces neuronal death after OGD and glutamate excitotoxicity, respectively, whereas treatment of neurons with FK866 increases neuronal death after OGD. Furthermore, over-expression of hNAMPT reduces glutamate excitotoxicity, whereas over-expression of hNAMPT mutants lacking enzymatic activity exerts no effect on neuronal death. In addition, the addition of NAD⁺ and NAM increases mitochondrial biogenesis in neurons after OGD. Over-expression of NAMPT in neurons reduces mitochondrial membrane potential depolarization following glutamate stimulation, whereas over-expression of hNAMPT mutants without enzymatic activity does not affect mitochondrial membrane potential depolarization. Thus, NAMPT exerts neuroprotective effects in ischemia via enzymatic activity for NAD⁺ production and amelioration of mitochondrial dysfunction [84].

Zhao et al demonstrated that plasma concentrations of NAMPT, NAD⁺, and ATP were all increased in mice after cerebral ischemia [83]. For the first time, they reported that cultured glia, but not neurons, increased their secretion of NAMPT under OGD stress conditions, as shown by immunoblotting and enzyme immunoassays. Treatment with wild-type NAMPT, but not H247A-mutant enzymatic-dead NAMPT, significantly attenuated OGD-induced cell death and apoptosis in both cultured mouse neurons and glia. Conversely, treatment with neutralizing antibodies abolished the protective effects of eNAMPT on cell viability, although this treatment failed to block the anti-apoptotic effects of eNAMPT [83]. Furthermore, plasma NAMPT concentrations were decreased in 6-month-old but not 3-month-old spontaneously hypertensive stroke-prone (SHR-SP) rats, a spontaneous stroke animal model, compared with age-matched Wistar-Kyoto rats. The decrease in plasma visfatin levels in 6-month-old SHR-SP rats is correlated with the onset of susceptibility to stroke at around 6-months of age [85]. More importantly, treatment with FK866 for 3 months does not induce death or abnormal behavior in normal rats, although inhibition of NAMPT enzymatic function by FK866 accelerates stroke occurrence and death in SHR-SP rats [83]. The authors of these studies argue that eNAMPT exhibits enzymatic activity and may be as neuroprotective as iNAMPT in cerebral ischemic injury.

Cerebral ischemic injury leads to both gray matter and WM pathology. WM is highly vulnerable to ischemia and often more severely injured than gray matter [86]. In most cases of human stroke, WM injury accounts for half of the lesion volume [87]. Despite the essential roles of WM in neural function and the susceptibility of WM to ischemia, WM injury has been largely overlooked in animal studies as well as in clinical treatments [88]. However, complete neuroprotection cannot be attained without WM protection. NAMPT can delay axon degeneration in the presence of NAM in an in vitro Wallerian degeneration assay [89]. Sasaki et al. suggest that stimulation of NAD biosynthetic pathways by a variety of interventions may be useful in preventing or delaying axonal degeneration [89]. Our group studied the role of NAMPT in both gray and white matter protection against ischemia in vivo. We confirmed the protective effects of NAMPT against ischemic stroke induced in mice by transient occlusion of the middle cerebral artery. Neuron-specific NAMPT overexpression significantly reduced infarct volume and improved long-term neurologic outcomes compared with littermates. Interestingly, neuronal overexpression of NAMPT increased the area of myelinated fibers in the striatum and corpus callosum, indicating that NAMPT protects against WM injury. We then provided evidence that the protective effects are mediated by eNAMPT. Conditioned medium from NAMPT-overexpressing neurons exposed to OGD protected cultured oligodendrocytes from OGD and these effects were abolished with neutralizing antibodies [18]. Thus, we concluded that NAMPT may not function simply in a restricted intracellular manner during stroke recovery but that it protects both gray matter and WM from stroke injury through extracellular actions.

In sum, most studies convincingly support a protective role for iNAMPT against brain injury. However, there is less agreement on the effects of eNAMPT on ischemic injury. In contrast with the studies mentioned above, which indicate protective roles of NAMPT against cerebral ischemic injury, Zhao et al reported that cerebral ischemic-like injury in cultured neural cells was exacerbated by eNAMPT. Furthermore, administration of recombinant NAMPT protein exacerbated MCAO-induced neuronal injury in rat brain. NAMPT exacerbated OGD-induced neuronal injury only in mixed neuron-glia cultures, but not in pure neuron cultures. In the mixed cultures, NAMPT protein promoted TNF-α release in a time- and concentration-dependent fashion. Conversely TNF-α neutralizing antibodies protected against the enhancement of OGD neuronal injury by NAMPT. In addition, the H247A mutant of NAMPT lacking all enzymatic activity exerted similar effects on ischemic neuronal injury and TNF-α release as the wild type protein. Thus, the authors concluded that eNAMPT is an injurious and inflammatory factor in cerebral ischemia and aggravates ischemic neuronal injury by triggering TNF-α release from glial cells, through a mechanism unrelated to NAMPT enzymatic activity [8].

8. More considerations THAT may affect the role of NAMPT in ischemic brain

The precise role of NAMPT in cerebrovascular diseases, including ischemic brain injury, needs to be fully elucidated because of its controversial biological actions on a variety of cellular functions. Ischemic stroke damages many different types of cells in the brain aside from neurons and glia. For example, dysfunction in macrophages, vascular smooth muscle cells (VSMCs), and endothelial cells is believed to to underlie the genesis and development of atherosclerosis, and it is generally accepted that NAMPT is a proinflammatory factor in atherosclerosis. However, NAMPT is not only an inflammatory cytokine that promotes endothelial cell activation, VSMCs proliferation, and macrophage migration, but also a key protector of the same cells against various stresses [37,40,54,90,91]. These complex features of NAMPT are consistent with the evolution of cytokines as part of an adaptive immune response to injury. Inflammatory mediators affect the integrity of the neurovascular unit, including the endothelium, astrocytes, the vascular extracellular matrix, pericytes, and neurons and their axons. Indeed, all components of the neurovascular unit respond to inflammation, which is known to be initiated by ischemia. Under ischemic conditions, pericytes constrict brain capillaries and then die, which may lead to a long-lasting decrease in blood flow and loss of blood–brain barrier function, thereby increasing death of nerve cells [92]. However, how other components of the neurovascular unit communicate during post-ischemic inflammatory responses is not yet clear, as a unifying scheme of inflammatory responses has not yet been developed. As such, the effects of NAMPT on cellular processes may be double-edged and depend on its expression level and distribution. Thus, many issues remain unsolved, although studies in various tissues have provided valuable information on the relationship between NAMPT and ischemic stroke and improved our understanding of the underlying mechanisms.

9. Concluding Remarks

NAMPT has drawn much attention in biomedical fields because of its pleiotropic physiological functions and because its dysregulation is implicated in a number of human diseases, including ischemia. Recent studies have provided new insights into the molecular effects of NAMPT using the specific inhibitor FK866 and genetic means of manipulating NAMPT expression. Collectively, the majority of experimental studies conducted thus far suggest that NAMPT is an attractive therapeutic target for both the prevention and treatment of ischemic stroke. Zhao et al have commented on some discrepancies between studies, suggesting that the duration of OGD may be critical [83]. In addition, standardized procedures for the preparation of NAMPT protein constructs are urgently needed. Further exploration is needed of the potential effects of trace amounts of residual endotoxins from E.Coli in recombinant preparations of NAMPT. Moreover, some reports have claimed that NAMPT is a ligand for the insulin receptor [28,58]. However, the actual receptor for NAMPT has not yet been convincingly identified. The mechanisms underlying the cytokine-like effects of NAMPT also await further exploration. It is worth noting that the extracellular mode of action of NAMPT and its ability to cross the blood-brain barrier both make it a clinically viable candidate for stroke therapy. In conclusion, further investigations into the properties of NAMPT are highly warranted.